6v Battery Low Voltage Monitor

Hi,

The way it would work with just a couple caps and resistor addition would be as follows...

As the voltage supply came down lower and lower, eventually it would reach the point where the LED starts to flash. The flashing would be very short duration at first, but then get longer and longer. Thus, it would start out very narrow but then the LED would flash brighter and brighter until it reached the max duration which would probably be around 50 percent duty cycle. That could mean one for 2 seconds and off for 2 seconds as the supply got lower.

That only takes a couple caps and a resistor, but maybe you would prefer to have a more constant LED blink rate and 'on' time. It's up to you here. We could easily add a couple components like transistor or something to get any operation really.

MrAl said:

We could easily add a couple components like transistor or something to get any operation really.

Click to expand...

I think I have the current circuit correct in the attached gif, if so that should be actual size of the circuit to the right, if I mount the resistors radially and use radial capacitors on the .1" grid board. The 10K value on the trimmer is the same as the trimmer used for the LCD contrast, and I bought spares.

Looks like there's room for a T0-92 transistor for a more "noticeable" indication. I'm figuring some if not all of the R/C components (and the LED) would move from the LM431 to the collector of a NPN transistor, and the transistor's base connected to where the diode is now?

Hi,

That looks right so far. You want to try it that way first and see if you like the results?

MrAl said:

Hi,

That looks right so far. You want to try it that way first and see if you like the results?

Click to expand...

On a purely economic level, I'd prefer to buy components I may not use on this project (but may on another down the road) than to pay 10X their costs for shipping in an additional order a week later (I currently don't have a project that overlaps this one to "cushion the blow").

If you don't mind posting a schematic of your thoughts on the "advanced" version I'd appreciate it, since I wouldn't mind using "fresh" components for everything in this project. Thanks!

What about using a "Flashing LED" - - saves any additional circuitry . . .

MrAl said:

Hi,


Here's a simple circuit to monitor under voltage.
R3 sets the LED 'on' current.

Click to expand...

Thank you MrAl. This is exactly what I was looking for.

MrAl said:

Hi Roger,

You brought up a good point about the dieing battery. The OP wanted a circuit to turn the LED 'on' when the voltage goes under a certain limit point, but the more usual under voltage monitor turns the LED 'off' when the voltage goes too low. This logic might be a bit better because if the battery goes really really low then the LED wont turn on anymore anyway. To get the LED to turn 'off' with low voltage the LED is placed in series with the LM431 (cathode) instead of in parallel with the LM431. The resistor is still sized for the LED.

Click to expand...

That's because my main load is a really powerful lamp (H3, 6V 55W). I want to cut the lamp off when battery voltage falls, and light this red LED on. The LED just alerts.

ozarkshermit said:

What about using a "Flashing LED" - - saves any additional circuitry . . .

Click to expand...

Already considered, but my project (which might be better moved to it's own subject, but it started here) is using four LTL-433 2x5mm rectangular LEDs that fit flush against the side of a HD44780 LCD display. It's an aesthetic decision, but one I'd like to keep. If there's a flashing rectangular in that size I'm unaware -- but let me know if you have a source. All I could find are round.

Hi again,

Ok Max, if you like the way it works as is i guess you are all set then

Roger:
It is starting to look like the best bet is to follow the circuit we had to start with (with the LM431) and go with a 555 timer ic following that. The reason being twofold: One, to get more reliable and repeatable timing for the LED, and Two, to allow the use of much smaller capacitors. The other version was starting to required two caps with values of 500uf or better but using a CMOS 555 timer we can use much smaller caps. The LED flashing will be very controllable via a couple of resistors.
Sound good to you?

MrAl said:

Hi again,

It is starting to look like the best bet is to follow the circuit we had to start with (with the LM431) and go with a 555 timer ic following that. <snip>

Sound good to you?

Click to expand...

Give it a whirl. I have NE555s on hand and it may keep me from having to remove the 8 pin socket already soldered on the protoboard the failed LM10 attempt was using.

The 555 pulls milli rather than micro amps, so I may end up using the original circuit after I breadboard the new one and hook up a meter. But I really appreciate you going to the extra effort.

Hi Roger,

Here's one way to do it. If you are worried about the current draw of a regular 555 then maybe we can modify this circuit a little to turn the 555 off completely when the voltage supply is in spec. That way it would only be on during the LED pulsing.

This circuit pulses the LED for 0.5 seconds every 5 seconds approximately. Changing the resistors R1 and R2 modify both of these timings. R1 for 'on' time, R2 for 'off' time.

MrAl said:

Hi Roger,

Here's one way to do it. If you are worried about the current draw of a regular 555 then maybe we can modify this circuit a little to turn the 555 off completely when the voltage supply is in spec. That way it would only be on during the LED pulsing.

Click to expand...

First off, my thanks to the creator of this forum, and a BIG thanks to you, MrAl, for taking your time to personally help the folks who wander in and out of here.

I've got a 555 breadboarded, using a 22K for R1 and a 22uf for C1, running astable (without R4, R5 and Q1). The circuit is connected to 6 AA alkaline batteries (used ones, without the circuit load I show around 7.4V, with 6.5V). Current draw is around 4.2mA off, 6.7mA on.

This circuit pulses the LED for 0.5 seconds every 5 seconds approximately. Changing the resistors R1 and R2 modify both of these timings. R1 for 'on' time, R2 for 'off' time.

Click to expand...

I made some changes to R1 to shorten the pulse duration (down to 6.8K) but it didn't seem to change the amperage draw enough to worry about. The timing cycle you have is perfectly adequate. The ALD7555PAL specs in the uA range looked promising, but the 10V max supply voltage killed that idea due to my 12V automotive power option. I guess I'll just stick with a NE555N.

Realistically the Tiny Trak4 (TT4) documentation says it draws an average of 40mA, the LCD specs are 2mA to 5mA without backlight (72mA to 80mA with, which is why I included the option to turn it off), I'm not sure of the additional requirements for the keyboard (5mA to 10mA assumed), but the keyboard wouldn't be attached all the time. So a worst-case 150mA draw on a 2000mAH battery pack could be in the neighborhood of 12 hours if my math is correct. The devices to be attached to the TT4 are a Garmin GPS12 that claims 24 hrs operation with 4 AA alkalines (actual reviews say bring spares if you plan to use it all day) and an Icom IC-T90A transceiver that specs 5 hrs at 1x8 TX/RX duty cycle at full 5W power, I'm not sure if I can get 12 hours out if it at the lower 1W setting. Add to that the TT4 would be programmed for "Smart Beaconing", which would key the push-to-talk on the radio for a lat/long transmission depending on speed and direction changes, depleting the radio's battery more than just a simple time interval setting would if I move around a lot. So in practice, the "Low-Battery" circuit indicator may never come on before the batteries in the peripherals die, but it will be interesting to find out.

I'll order up the LM431 and fresh passives and report back later in the week on what I end up doing. Thanks so much!

Hi again,

Hey you're welcome. I'd like to hear more about this too once you get it built up and start testing it out.

MrAl said:

Hi again,

Hey you're welcome. I'd like to hear more about this too once you get it built up and start testing it out.

Click to expand...

Well then , here we go.

Fully charged, the 7.2V 2000mAH NiMH battery pack read 8.42V. I put four 1N4003 diodes in series so their .6V voltage drops gave me the observed steps 8.42V, 7.70V, 7.10V, 6.45V and 5.80V. I built up Circuit #1 (I didn't have the originally specified 750 ohm LED current limiting resistor and used a 681 ohm instead) and set the 10K trimmer so that the LED began illuminating at the 7.10V reading. The circuit worked as you described, the LED turned off when I went above the 7.10V point and remained lit when I went below. Amperage readings were:

8.42V - 9.68mA LED off
7.70V - 8.69mA LED off
7.10V - 6.82mA LED on
6.45V - 5.89mA LED on
5.80V - 4.98mA LED on

I added the 10K resistor plus the 10uf and 1uf caps in Circuit #2 and there was no change in operation - no dim illumination or flashing. The mA readings were the same as with Circuit #1. I checked everything for continuity, all looked good, so I shrugged my shoulders and pressed on.

I pulled the two caps, diode and LED, moved the 10K, added the 20K and the PNP, connecting its collector to the trigger input of the 555 timer circuit that had been working as an astable and built Circuit #3. There was no LED illumination at any voltage. I set the voltage at 7.10V and tweaked the 10K trimmer, but there was no change. I removed the collector output from the 555's trigger and powered the 555 as an astable and it still worked, I replaced the PNP, hooked everything back up and rechecked connections but the LED still would not illuminate.

I found it interesting that with Circuit #1 the current draw actually went down with the voltage, enough so that I'm bringing a second DMM from home tomorrow. If nothing else I can run concurrent V/A checks. I worked from my attached drawing, if you see anything I missed from your originals, let me know.

Thanks!

Hi MrAl.

I brought in the second DMM and Circuit #1 had similar readings with it. Being a student of the Forrest M. Mims III "Mini-Notebooks" from Radio Shack that always interfaced with a NPN, I decided to try one between the 555 from the LM431 now that I saw where you were taking the transistor bias for you PNP switch. Rather than use the 555's pin 2 trigger input, I switched Vcc to the 555 (already working fine as an astable) instead. I also changed the 681 ohm LED current limiting resistor to 1K. See the attachment, which includes the schematics and a chart of the voltages and the current draw of the two circuits tested. The battery's voltage had dropped slightly overnight, and in addition the current readings read lower with the Tenma 72-4025 used today than yesterday's Radio Shack 22-178 (now being used exclusively for voltage readings).

Since the 555 is off when voltage is 7.06 and above, current is almost identical. When the 555 is switched on but the LED is not on, current is around 500uA less. When the LED blinks, the current is about the same as when the LED is on without the 555 addition.

I'm debating if the 500uA saved during the 555's timing cycle is worth the extra PCB real estate the additional components take up. I'll probably go with the lower parts count, but it was fun getting to that conclusion.

Thanks for the "heads-up" on the LM431 and your time on everything!

Hi Roger,

Im glad you got it working. So are you happy with the blink rate and the way it works overall?
Were you saying the circuit with only one transistor didnt work right?

MrAl said:

Hi Roger,

Im glad you got it working. So are you happy with the blink rate and the way it works overall?
Were you saying the circuit with only one transistor didnt work right?

Click to expand...

Correct.

Your original LM431 w/voltage divider circuit from Post #9 works.

Adding the R/C components described in Post #21 which I drew in Post #22 did not work. I did not check it with a logic probe or scope to see if it could actually be pulsing, but too fast for the human eye to recognize.

The 555 circuit alone (without the PNP transistor, R4 & R5, the LM431 and voltage divider) in Post #30 works as a simple astable LED blinker. I -did not- get your entire circuit to work. The blink rate was fine.

Post #33 contains a thumbnail of all three circuits, I breadboarded from a printed copy of that.

Post #34 is something I came up with that uses a NPN rather than PNP transistor to interface the two elements that I did get to work, which were the original LM431 circuit from Post #9 and the astable section of the 555 timer from Post #30.

RogerE said:

Correct.

Your original LM431 w/voltage divider circuit from Post #9 works.

Adding the R/C components described in Post #21 which I drew in Post #22 did not work. I did not check it with a logic probe or scope to see if it could actually be pulsing, but too fast for the human eye to recognize.

The 555 circuit alone (without the PNP transistor, R4 & R5, the LM431 and voltage divider) in Post #30 works as a simple astable LED blinker. I -did not- get your entire circuit to work. The blink rate was fine.

Post #33 contains a thumbnail of all three circuits, I breadboarded from a printed copy of that.

Post #34 is something I came up with that uses a NPN rather than PNP transistor to interface the two elements that I did get to work, which were the original LM431 circuit from Post #9 and the astable section of the 555 timer from Post #30.

Click to expand...

Hi,

Yes with the added RC components it must have been blinking too fast. We'd have to increase the capacitances.

MrAl said:

Hi,

Yes with the added RC components it must have been blinking too fast. We'd have to increase the capacitances.

Click to expand...

Yeah, but in the case of the 555, in the final analysis additional components to blink the LED didn't offer appreciable savings in current, so why bother?

Hi,

Did you try a larger cathode resistance? That might reduce the current drain. The idea is to look for the major cause of the current drain and do something about it which usually means increasing a resistance. The LM431 i think can be set up to draw less current, so the remaining current draw is from the 555 which after cutting the power to should mean a less overall current draw. IF that doesnt work, then it means the LM431 might be set up to draw too much current already. Take a look and see what you can find.

MrAl said:

Did you try a larger cathode resistance?

Click to expand...

The original 750 ohms was breadboarded with a 681 ohm, I changed that to a 1K when I built the NPN circuit. Those were only small changes in value, so I'd really have to say that I didn't pursue the matter. Since I omitted the TT4's blue "power on" LED (the active LCD display would serve as a "power on" indicator anyway), I figure the current saved by that could possibly be equal to that used by the LM431 "low-battery" circuit, so everything could possibly even out.

I've got the project soldered up, installed in the case and ready for programming:

https://www.electro-tech-online.com/custompdfs/2012/05/TT4-DK_Construction.pdf

Now my concern is the "flexible, roll up keyboard" I bought off eBay is USB with a PS/2 adapter, and doesn't work with the project. A regular, rigid PS/2 keyboard does work with the project, and the flexible one with the adapter does in a PS/2 port on a PC. Looks like my next move is to cut the USB controller section open on flexible one and see if the switch matrix is compatible with a PS/2 controller board out of a standard rigid keyboard.